Health expenditures and insurance plans
Data from the RAND Health Insurance Experiment. The data comes from Deb and Trivedi (2002). It includes variables on the number of contacts with a medical doctor, medical expenditures, demographics, health status, and insurance status. Cameron and Trivedi (2005).
data
Format
A data frame with 20,190 observations and 45 variables:
- plan: health insurance plan number
- site: one of six sites where experiment was conducted
- coins: medical coinsurance
- tookphys: took baseline physical
- year: study year
- zper: person id, leading digit is sit
- black: = 1 if race of household head is black
- income: income based on annual income
- xage: age that year
- female: = 1 if person is female
- educdec: years of schooling of decision maker
- time: time eligible during the year
- outpdol: outpatient exp. excl. ment and
- drugdol: drugs purchased, outpatient
- suppdol: supplies purchased, outpatient
- mentdol: psychotherapy exp., outpatient
- inpdol: inpatient exp., facilities and md
- meddol: annual medical expenditures in constant dollars, excluding dental and outpatient mental
- totadm: number of hospital admissions
- inpmis: missing any inpatient charges
- mentvis: number psychotehrapy visits
- mdvis: number face-to-face md visits
- notmdvis: number face-to-face, not md visits
- num: family size
- mhi: mental health index, baseline
- disea: number of chronic diseases
- physlm: = 1 if person has physical limitation
- ghindx: general health index, baseline
- mdeoff: maximum expenditure offer
- pioff: participation incentive
- child: = 1 if age is less than 18
- fchild: = 1 if female child
- lfam: log of family size
- lpi: log of annual participation incentive payment or 0 if no payment
- idp: = 1 if individual deductible plan
- logc: log(coinsurance + 1) where coinsurance rate is 0 to 100
- fmde: log(max(medical deductible expenditure)) if idp=1 and mde>1, 0 otherwise
- hlthg: = 1 if self-rated health is good
- hlthf: = 1 if self-rated health is fair
- hlthp: = 1 if self-rated health is poor, (omitted is excellent)
- xghindx: ghi with imputation
- linc: log of annual family income, usd
- lnum: log of family size
- lnmeddol: log of medical expenditures given meddol > 0; missing otherwise
- binexp: = 1 if medical expenditures > 0
Source
http://cameron.econ.ucdavis.edu/mmabook/mmadata.html
randhealth
Section in Text
16.6 Selection Models, pp. 553-6, 565 20.3 Count Example: Contacts with Medical Doctor, p.671
Examples
summary(randhealth)
References
Cameron, A. and Trivedi, P. (2005), "Microeconometrics: Methods and Applications," Cambridge University Press, New York.
Deb, P. and Trivedi, P.K. (2002), "The Structure of Demand for Health Care: Latent Class versus Two-Part Models," Journal of Health Economics, 21, 601-625.
RAND Corporation. "RAND's Health Insurance Experiment ." https://www.rand.org/health-care/projects/hie.html